In the prior art, when a user uses a mobile terminal to send a short message or an email to a recipient, he/she usually utilizes an existing input method to edit the content of the short message or email. Since the input method itself has an intelligent learning function and can learn the user's input habit and frequently-used words, the editing speed can be improved to some degree when the user edits the content of a short message or an email.
However, the intelligent learning function of existing input methods is only designed for the input method itself, which is to say, the input method cannot carry out intelligent learning on the basis of the recipient when the user edits the content of a short message or an email; consequently, the user still has to input the content of a short message or an email with an existing input method even when the user wants to send a short message or an email to a recipient who is in frequent contact with him/her. This drawback results in inconvenience when the user wants to rapidly send a short message or an email to a recipient who is in frequent contact with him/her.